Market Pulse Universal

Specifications

Release: May 1, 2024

Version: 1.0.0

Size: >500G parquet

Update Freq: Daily

Usage: This is a daily file of buyer intent, with both B2B and B2C topic areas. The file contains 6-10 billion signals, normalized to a hem/topic pair. There are two files in the delivery: a daily data file, and topic taxonomy lookup table. All intent records have a hashed email (sha256_lc_hem), and on average, 42% of the records are also linked to a up_id.

Market Pulse Universal Schema

Feature	Description	Sample
sha256_lc_hem	The most observed Sha256 lowercase Hashed EMail (HEM) for this person. Used for identity management and file matching.	e81696563daa221a 9972c59d0a7b3728d 81b2c3b5f95cfaa2 c425426a50a48e4
id	This is the universal person identifier, up_id.	5x5-7261f2df14aa7a153013de44d9ce3b53
topic	This is a persistent alpha numeric id, unique to each topic.	4eyes_105491
score	This score represents the strength of the observed signal for the listed topic, based partly on the frequency of observatios, and partly on the Large Language Model precision of how closely the page visited matches the listed topic.	medium

Taxonomy Lookup Table

Feature	Description	Sample
id	This is the topic identifier, and may be joined to the 'topic' field in the daily data file.	100006
category	This is the top level category for the listed topic, there are currently 26 categories.	Apparel & Accessories
sub_category	The is the one-level drill down from the category for the listed topic. There are currently 364 subcategories.	High End
topic	The is the english language topic. There are currently 21,467 unique topics, covering both B2B and B2C intent.	Armani